ghost peaks, small standard = inaccuratly high analytes

[Jessamaka]

I'm doing quantification of VFAs (acetic, propionic, isobutyric and butyric) in paper in aqueous solutions. The samples are acidified with HCl and internal standards are either crotonic or pivalic acid.
My parameters are:
Focus GC FID
Column; Stabilwax-DA column (Restek 30m-0.25mm ID-0.50μm df)
Four Temp; 100°C to 190°C @ 5°C/min, 240°C @10°C/min
Inj. temp; 200°C split 7:1
Carrier; N2 constant pressure 225kPa
Det. Temp; 200°C

The problems are these, ghost peaks on blank injections and decreased standard peak area for samples as compared to calibration injections.

Problem #1 a blank of water has peaks for all the analytes and the standard. running with machine without a physical injection shows no peaks. The ghost peaks seem to be a problem of residue somewhere, but where? If I replace the septum, liner, syringe, solvent vials and solvent, it goes away. But with just one injection (10-100ppm analyte) they're back. It is not economical for me to replace all this each injection. I did some calculations and my vapor volume is 1/3 that of the liner volume so I'm doubting backflash is causing the residue. I'm rinsing the syringe 5 times with water and 5 times with acetone/methanol before and after injection and three times with sample. The peaks will minimize if I inject water many times, but who has the time for that? Ideally I'd like the analytes and standard to stop sticking to whatever they're sticking to.

Problem #2 I can run several calibration standards and get a pretty decent calibration using the internal standard method. The peak area is pretty stable. Then I inject a sample and the standard is a fraction of what it should be, giving falsely large values for analyte (the actual concentration is confirmed via IC). I should note this is not always the case. Depending on where the sample is from there may be no issues, I do the injection and the results are reasonable. But for certain others, if the results were accurate, I should be able to smell it across the room.
I've recently switched from crotonic acid to pivalic acid and the problem continues.
I know GC FID, water, and VFAs should never really go together, but this machine was purchased before I came to the company, and we'd all rather 'fix' these problems as best as possible instead of returning to paying someone else to do them on an IC.

Thanks for the insight,
Jessica

[Don_Hilton]

Increased flow out the split vent and increased temperature in the inlet may help. You are looking in the right places for carry over. The syringe can be an issue because acids are not terribly soluable in acidic water. And a water wash may not be particularly helpful in cleaning the syringe.

Make an injection and try just changing the syringe and make an injection with a clean syringe. If the carryover goes away, then you have nailed it down to one item. And if it is the syringe, I would wash first with acetone and then with the best solvent you can find for the organic acids.

The inlet is a possibility also. I would look for a way to increase the purge flow after injection. With a 7:1 split, what about the idea of making a splitless injection and after the purge off time, increasing the flow up to something like 50:1? (Or if you are using an instrument with gas saver, use that to kick up the flow rate sometime after injection).

And not that you asked, injections of acidified water are a bit rough on a wax type column. I would think that an extraction into something like dichlromethane might help - allow for injection in a well behaved solvent, and be a good wash solution in the GC as well. (I expect that we are looking at a slurry of paper in water, but if you can press out enough water to extract... an idea...)

[chromatographer1]

Don's comments are well taken.

The problem with the method as you are doing it is that, beside the locations already discussed, any residue in the injector can be a source of retaining the free acids. And residue can come from any salt mineral, or heavy organic left after the vaporization of the sample, especially water samples.

Water can even react with the glass producing hydroxyls which are then reactive sites for the free acid organics. Final effect? Carryover and ghosts peaks !

Using methanol as a solvent instead of water can be helpful especially if you use a plug of it behind the sample in the syringe and perform a slow injection (Varian and some other manufacturers offer this option) which can be done manually with a capable analyst.

DCM acetone and formic acid can also be used, formic acid also 'blocking' or occupying the active sites in your injector and the head of your column. You can even do an injection of formic acid immediately prior to your sample injection. Formic acid has little response with an FID. Still, wait for the main peak of it to pass before injecting your sample.

But as this problem of free acids by GC has been discussed thoroughly in other threads on this forum, if accuracy and reproducibility are important factors in your testing and it is worth it to you to have correct numbers, use IC or LC to do your analysis.

best wishes,

Rodney George
consultant

[Peter Apps]

Hi Jessica

Rod and Don have pretty much covered it (as usual ). The most likely place for carryover to come from is the inlet and you have two conflicting requirements - you need the minimum surface area to reduce adsorption, but you need an increased area to enable the water to evaporate. Try putting in a pre-injection dwell time of 3-5 s on the autosampler (if you are using one) - this will allow the needle to heat up by conduction from the inlet, and may improve vaporization. You may need to play a bit with dwell times and injection speeds to find an optimum.

If you have a programmable inlet then increasing both flow and split ratio after injection might help to flush out carryover.

Although it is very unlikely to be the cause of the problem - if you are programming your column to 240C your detector needs to be at least at 250C to stop heavy crud condensing into it.

Peter

[BZK]

Water require an high vaporizing heat. After injection vaporizing it isn't immediate and could be incomplete with a small doplet of water and vapour on the same time. Any cold spot such as stainless tube of injector (carrier line) can cause condensation, particulary with water as solvent.
This is, from my point of vue, your main problem.

Robertino Barcarolo